UK Firms Offshoring AI Workloads Due to High Energy Costs
New report reveals power constraints and costs are driving AI infrastructure out of Britain
One in five UK organizations have already moved their AI workloads overseas due to high electricity prices and grid constraints. A new industry report warns that rising energy costs are becoming a primary bottleneck for the UK's ambitions to become a global leader in artificial intelligence.
Key details
The "Land, Power, Compute" report, published by CUDO Compute, highlights a growing trend of "AI offshoring" driven by the stark reality of energy economics. According to a survey of 700 senior AI decision-makers, a third of UK organizations state that energy costs are actively limiting their ability to scale.
For more than 20% of respondents, energy bills now account for more than one-third of their entire AI infrastructure budget. This pressure is particularly acute among "AI-first" companies, where 32% are considering moving workloads abroad to secure more affordable or reliable power.
The problem is not unique to the UK. In Santa Clara, California—the heart of the global chip industry—nearly 100 MW of newly constructed data center capacity is reportedly sitting empty, awaiting power grid connections that may not be available for several years.
Why this matters
This shift indicates that infrastructure availability—specifically power and cooling—is now a more significant driver of AI deployment costs than the price of the GPUs themselves. When energy costs become prohibitive, companies are forced to choose between sovereign compute and economic viability.
Context
The UK has some of the highest industrial electricity prices in the developed world, often tied to volatile gas prices. While the government has proposed measures to decouple electricity prices from gas and incentivize fixed contracts, the physical infrastructure of the grid continues to fall behind the rapid demand surge from generative AI scaling.
What happens next
The UK government is under pressure to accelerate grid connection timelines and streamline planning for data centers. In the interim, established data center hubs are likely to lose investment to regions with spare grid capacity and lower-cost renewable energy, such as the Nordics or parts of North America.
Source: The Register Published on AI Usage Global, author: AUG Bot
